NEW YORK – Researchers from Stanford University and the Chan Zuckerberg Biohub have developed a microbiome-focused metabolomics pipeline to characterize microorganisms in the gut, as well as interactions between those microorganisms and their host.
In a paper published in Nature on Wednesday, the researchers described their construction of an integrated mass spectrometry pipeline to accelerate the identification of microbiota-dependent metabolites in diverse sample types. Using a library of 833 metabolites, they created a reference dataset of metabolomic profiles for individual bacterial strains to enable multiple modes of analysis and discovery. In total, they analyzed 178 individual prevalent human gut microorganisms representing 130 species and spanning six phyla.
The researchers subsequently established deviations in the relationships between phylogeny and metabolism. Moreover, they used machine learning to discover a previously undescribed type of metabolism in Bacteroides, and they revealed candidate biochemical pathways using comparative genomics.
They also found that microbiota-dependent metabolites can be detected in diverse biological fluids from gnotobiotic and conventionally colonized mice and then be traced back to the corresponding metabolomic profiles of cultured bacteria. Their datasets are publicly available on GitHub.
Among other findings, they observed unique high producers or consumers of specific metabolites within their strain collection. For example, they found that Enterococcus faecalis and Enterococcus faecium produced high levels of tyramine, a biogenic amine known to modulate host neurological functions. By contrast, Clostridium cadaveris consumed high levels of pantothenic acid, a molecule that is associated with inflammatory bowel diseases.
"This large-scale in vitro screen enables us to identify numerous high-abundance, variably conserved, microbially derived metabolites that can be tracked in vitro and in vivo," the authors wrote.
They were also able to analyze large-scale relationships between strain metabolism, known as metabolonomy, and phylogeny. One of their findings was two groups of phylogenetically distant strains in two phyla, Firmicutes and Actinobacteria, that accumulated high levels of ornithine and citrulline in the absence of substantial downstream polyamine production. They performed comparative genomics starting with the ornithine-producing arc genes in Lactococcus lactis and found them to be conserved among the ornithine-accumulating strains, such as the Lactobacillales. Importantly, these genes weren't detectable in the non-ornithine-accumulating phylogenetic neighbors in both Lactobacillales and Actinobacteria, illustrating that orthologous gene-metabolite relationships may be preserved when metabolic phenotypes are separated from phylogeny.
Overall, they said, this customizable and expandable method of constructing a chemical standard library-informed metabolomics pipeline is tailored to detecting products of gut anaerobic biochemistry. The bacterial strain profiles it produced for this study demonstrate that substantial metabolic variation is common even between closely related strains.
"Our findings, along with emerging studies on microbiome-focused metabolomics and gut microbial metabolism, reinforce the limits of phylogeny or genome-scale analysis to provide direct measurement or prediction of metabolic phenotypes and the molecules that link the microbiota to host physiology," the authors added. "Our existing strain-specific genome-by-metabolic profile data provides a rich resource for the comparative discovery of genes and pathways that underlie bacterial phenotypic variation."